2024-25 team win projection contest

[TeemoTeejay]

I’ll do a joint victory with dtka like in the 2020 dunk contest Wont be allowed to enter competition after this I think either way for awhile lol

[Mike G]

Pulling away?

Code: Select all

.  avg err   rmse   r^2       .  avg err   rmse   r^2
TmTj   7.27   9.6   .46      KPel   8.14    9.6   .58
dtka   7.44   9.5   .46      24pr   8.14   10.6   .33
avgA   7.48   9.8   .42      Crow   8.21   10.3   .41
medi   7.58  10.7   .38      vegas  8.25   10.4   .39
bpmW   7.70   9.6   .44      emin   8.31   10.5   .39
DQin   7.75  10.1   .43      24py   8.59   11.6   .33
eExp   8.00  10.4   .44      eWin   9.14   11.4   .22
WShr   8.08   9.9   .40      bmgm   9.25   12.1   .33
.                            perW   9.38   11.6   .20

https://www.basketball-reference.com/fr ... _prob.html

[Crow]

Might look at impact on final results if I regressed to mean by 0.5, 1.0 or 1.5 in general or in a specific way. Probably would help.


PERw in last place.

We know PER was trying to get the top "right". Generally high usage guys. Maybe it does for that limited purpose. Not well across all roles.

[Mike G]

Code: Select all

.  avg err   rmse   r^2       .  avg err   rmse   r^2
TmTj   7.45   9.8   .45      KPel   8.31    9.8   .58
dtka   7.64   9.7   .45      24pr   8.37   10.8   .32
avgA   7.72  10.0   .41      WShr   8.39   10.2   .39
medi   7.79  10.9   .37      vegas  8.44   10.6   .39
DQin   7.88  10.3   .42      emin   8.48   10.8   .38
bpmW   8.03   9.8   .43      24py   8.66   11.7   .32
eExp   8.20  10.6   .43      bmgm   9.35   12.2   .32
Crow   8.22  10.4   .42      eWin   9.52   11.7   .21
.                            perW   9.79   11.9   .19

Code: Select all

e =   .24         e =   .87        e =   2.02       e =   6.00
4.98   medi      7.11   TmTj      9.73   dtka      13.5   KPel
5.37   DQin      7.33   dtka      9.85   KPel      14.3   dtka
5.37   TmTj      7.33   medi      9.85   TmTj      14.5   bpmW
5.69   dtka      7.50   DQin      9.86   bpmW      14.5   WShr
5.95   eExp      7.76   bpmW     10.21   WShr      15.4   DQin
6.07   Crow      7.84   eExp     10.30   DQin      15.9   Crow
6.36   emin      7.88   Crow     10.44   Crow      15.9   TmTj
6.39   bpmW      8.07   KPel     10.60  vegas      16.1  vegas

[knarsu3]

So seems like a bunch of people were interested in seeing where LEBRON/DARKO rank but don't see it included?

[TeemoTeejay]

So seems like a bunch of people were interested in seeing where LEBRON/DARKO rank but don't see it included?

From my understanding, Darko would be first, and using the BBI projections, EPM would be between AvgA and median, and LEBRON would be a bit behind BPM, going by the bbref projections which is whats been used so far

Paul george and nico harrison ruined me lmao

[Mike G]

Has anyone posted a link to these, with certainty that they were made pre-season?

FWIW, I have a blend of BPM, WS, PER, and eWins that would be in first place; but I didn't enter it before the season, and I've not listed it.

[TeemoTeejay]

Has anyone posted a link to these, with certainty that they were made pre-season?

FWIW, I have a blend of BPM, WS, PER, and eWins that would be in first place; but I didn't enter it before the season, and I've not listed it.

DARKO came out the day after or day of, the BBI ones are here https://www.bball-index.com/season-team-projections/

[Crow]

Both links were posted March 16th, page 12.
And Knarsu3 had his own set right below them.

The authors of these sets can report their final results, if needed.

1 day late in apparent publishing has not been a problem previously. Any see anything fishy?

Remember or caring to post here varies.

Several ways to look at results. Overall known / tracked winner, official entry winner, 2 other measuring sticks available if not content with average error...

[Crow]

Currently moved ahead of Vegas and Pelton. Breathing down the neck of ESPN experts.

Eighth to tied for 6th by the 3 measures of error, among the listed.

A few more days.

[Mike G]

Both links were posted March 16th, page 12.
And Knarsu3 had his own set right below them.

Aha, thanks. As the last post on the page, they never showed up for me; and it seemed people were still asking for them by p. 13.

It seems a slippery slope. Potentially unlimited 'other' sets of projections -- multiples from knarsu? -- coming in at all times of year?
If your guesses are looking good in March, submit them then? If not, fuggedaboudit.
I asked at least once about others around deadline time, and Crow got us KPel, vegas, and bmgm. No mas.

[TeemoTeejay]

I don’t think there’s an issue with including LEBRON or EPM projections since he went on here to post and it’s not like either would win anyway

[dtkavana]

include them all! the more, the merrier, imo. We could still. give TMTJ the contest victory while including the others in a separate table?

[Crow]

To win the "contest" you have to enter.

To "win" overall tracked, they haven't in past.

Vegas has "won" overall without entering. ESPN Experts won I think, with someone else entering them. That was an overall win. Same I think with Darko.

As said before, somebody calculate the final results of the affected non contest names if they want. Interpret results as want to.

Already 3 ways to "win" in some minds in the contest. With the outside lane it becomes 6. Or use in the contest / average and be done with it. People will say what they want.

There could be someone outside the tracked who does really well. Care or don't. Tout that if they want. Join officially in future or not. It is a bit loose.

[kmedved]

I posted the DARKO projections on Twitter, and they're available here: https://docs.google.com/spreadsheets/d/ ... =235611189), but I don't actually mind a standard that they need to posted/linked here to be included. It's fairly low cost/effort of us to do so, and I didn't do it. As noted, there are many many win projections out there, and including all of them seems unrealistic, so requiring a post here seems fine.

Also if I recall, I think my projections may have come in a few hours late this year. I promise I didn't use any "in-sample" data, but I would have missed the deadline this year regardless.